A Proportionate Robust Diffusion Recursive Least Exponential Hyperbolic Cosine Algorithm for Distributed Estimation

In this brief, a proportionate robust diffusion recursive least exponential hyperbolic cosine algorithm is proposed for distributed estimation. To be robust against impulsive noise, an exponential hyperbolic cosine function is utilized. The optimum closed form formula of diagonal gain matrix for the suggested proportionate version of the algorithm is suggested. Moreover, the mean stability of the proposed algorithm is provided theoretically. Finally, the simulation results show the efficacy of the presented algorithm in an adaptive network in comparison to some recent algorithms in the literature.